Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Use prescribed burning Terrestrial Mammal Conservation

Key messages

Read our guidance on Key messages before continuing

  • Thirty-seven studies evaluated the effects on mammals of using prescribed burning. Twenty-five studies were in the USA, three each were in Canada and South Africa, two each were in Spain and Tanzania and one each was in France and Auatralia.



  • Abundance (11 studies): Five of 10 replicated studies (of which eight were controlled and two were site comparisons), in the USA, Spain and Australia, found that prescribed burning did not increase abundances of small mammals. Three studies found mixed effects, on cottontail rabbits and small mammals and two found that burning increased numbers of European rabbits and small mammals. A systematic review in the USA found that two mammal species showed positive responses (abundance or reproduction) to prescribed burning while three showed no response.
  • Reproductive success (1 study): A before-and-after, site comparison study in South Africa found that 92% of Cape mountain zebra foals were produced in the three years post-fire compared to 8% in the three years pre-fire.
  • Condition (1 study): A replicated, controlled study, in the USA, found that prescribed burning did not reduce bot fly infestation rates among rodents and cottontail rabbits.
  • Occupancy/range (3 studies): Two of three studies (including two site comparisons and one controlled study), in the USA and Canada, found that prescribed burning resulted in larger areas being occupied by black-tailed prairie dog colonies and smaller individual home ranges of Mexican fox squirrels. The third study found that prescribed burning did not increase occupancy rates of beaver lodges.


  • Use (21 studies): Ten of 21 studies (including eight controlled studies and eight site comparisons with a further four being before-and-after studies), in the USA, Canada, South Africa, Tanzania and France, found that prescribed burning increased use of areas (measured either as time spent in areas or consumption of food resources) by bighorn sheep, mule deer, pronghorn antelope, elk, plains bison, Cape mountain zebrasand mouflon. Six studies found mixed effects, with responses differing among different ages or sexes of white-tailed deer, bison and elk, differing among different large herbivore species or varying over time for elk, while swift foxes denned more but did not hunt more in burned areas. The other five studies showed that prescribed burning did not increase use or herbivory by elk, black-tailed deer, white-tailed deer or mixed species groups of mammalian herbivores.
  • Behaviour change (1 study): A site comparison study in Tanzania found that vigilance of Thomson’s gazelles did not differ between those on burned and unburned areas.

Supporting evidence from individual studies


A site comparison in 1975–1978 of shrubland and grassland at a site in Idaho, USA (Peek et al. 1979) found that bighorn sheep Ovis canadensis consumed bluebunch wheatgrass Agropyron spicatum growing in burned areas more than they consumed on unburned areas. In the first summer after burning, a higher proportion of bluebunch wheatgrass stems was grazed on burned areas (73%) than on unburned areas (8%). The same pattern was observed, though with reducing magnitude, two years after burning (66 vs 25%), three years after burning (30 vs 10%) and four years after burning (36 vs 22%). Within an 86-km2 study area, seven areas (0.05–0.45 ha, total area 1.51 ha) had controlled burns carried out in September 1974. One hundred randomly selected bluebunch wheatgrass stems from burned and unburned areas were inspected each year to calculate the proportion that was grazed.


A replicated, controlled study in 1975–1977 on grassland in British Columbia, Canada (Willms et al. 1980) found that in burned areas, bluebunch wheatgrass Agropyron spicatum was consumed more by foraging mule deer Odocoileus hemionus than it was in unburned areas. Deer took more bites/observation of bluebunch wheatgrass in burned plots (average 22 bites) than in unburned plots (average two bites). Plots were studied at two sites in sagebrush and two in Douglas fir Pseudotsuga menziesii forest. At each site, plots (1.25 × 5 m) were established in a block. In each block, in October 1975, three plots were burned and three were not burned. In April 1976, three deer were fenced onto the block and their selection between plots was assessed through direct observations at intervals through the day. The same three deer were used on all blocks and observed twice/block for one day each time. In April 1977, four deer were observed, on two blocks combined, over four days.


A randomized, replicated, controlled study in 1971–1974 of a grassland in Washington, USA (Skovlin et al. 1983) found that burning grass did not increase overwinter use by Rocky Mountain elk Cervus canadensis nelsoni. Overwinter use by elk totalled 47–80 elk days/ha on burned areas and 42–79 elk days/ha on unburned areas. Within each of six plots, one 3.1-ha subplot was randomly assigned for burning and one was not burned. Burning was carried out once, in late-autumn 1971. Elk pellets were counted in spring to assess use of plots in the winters of 1971–1972, 1972–1973 and 1973–1974.


A replicated, controlled study in 1981–1983 of a pinyon-juniper woodland in New Mexico, USA (Severson 1986) found that felled forest areas that were burned did not have more small mammals than did felled unburned areas 13–18 years after treatment. A similar number of small mammals was caught in stands that were bulldozed and burned (408) as in stands that were bulldozed without burning (433). Fewer were caught in undisturbed stands (246). Treatment plots, c.120 ha each, were established in each of two woodland blocks, one in 1965, one in 1970. In each block, one plot was bulldozed (trees pushed over and left), one was bulldozed with trees pushed and piled, then burned and one was undisturbed. Small mammals were trapped in the second and third week of September, each year, in 1981–1983. Each plot was sampled for four days each year.


A replicated, site-comparison study in 1980 of forest in Alberta, Canada (Bentz & Woodard 1988) found that previously burned areas were used more by Rocky Mountain bighorn sheep Ovis canadensis canadensis than were unburned areas. In all five comparisons, at different distances below the treeline, more sheep pellets were found in burned areas (14–424 pellet groups/ha) than in unburned areas (0–108 pellet groups/ha). Three fire-modified sites (burned in 1919–1970) and three unburned sites (average forest age of 81–256 years old) were studied. At each site, three transects ran downslope from the treeline, to the valley bottom. Relative use by sheep of each area was assessed by counting pellet-groups in randomly located plots along these transects in 1980.


A site comparison study in 1985–1986 of prairie in Alberta, USA (Courtney 1988) found that pronghorn antelope Antilocapra americana made greater use of burned areas, relative to their availability, than of unburned areas in five of the 12 months surveyed. The number of pronghorn groups on burned areas was greater than expected in September, October, November, January and April. During these months, 5–22 pronghorn groups were found on burned areas, from totals of 38–97 pronghorn groups overall. If no preference was shown for or against burned ground, 5% of groups would be expected on it. Pronghorns especially favoured burns containing pricklypear cactus Opuntia polyacantha. Areas were burned in July–August 1985. Pronghorn were surveyed monthly, from July 1985 to June 1986. Groups <1 km away were mapped along a 138-km route, travelled by vehicle at <50 km/hour.


A site comparison study in 1983–1985 of a shrubland site in California, USA (Klinger et al. 1989) found that prescribed burning did not increase use of such areas by black-tailed deer Odocoileus hemionus, relative to unburned areas. There was no significant difference in density of faecal pellet groups between burned and unburned plots over the two years following burning (data not reported). In an area of chaparral shrubland, approximately 20% (7km2) was burned in November 1983. Twenty-three circular plots, each 100 m2, were surveyed for faecal clumps. Eleven plots were in burned areas and 12 were in unburned areas. Faecal pellet clumps were cleared and counted at end of each wet (November–April) and dry (May–October) season from November 1983 to October 1985.


A replicated, randomized, controlled study in 1983–1987 of a rough fescue Festuca scabrella-dominated grassland in Montana, USA (Jourdonnais & Bedunah 1990) found that burning increased grazing of rough fescue by elk Cervus canadensis nelsoni in the first, but not subsequent, winters following burning. The proportion of rough fescue plants grazed was greater in burned plots (39%) than in unburned plots (15%) over the winter following autumn burning. Over the subsequent three winters, the proportion grazed did not differ between burned plots (including spring burns; 98–100%) and unburned plots (87–97%). Additionally, a higher proportion of rough fescue biomass was utilized over the first two winters following burning (82–86%) than was utilized in unburned plots (24–69%). Six plots were burned on 17 October 1983 and six on 15 April 1984. Three plots were not burned. Plots were 2 ha in extent. Treatments were assigned randomly. Elk utilization of rough fescue was assessed by inspecting the closest plant to 50 points along each of two transects/plot and determining the proportion that was grazed. Additionally, six caged and six non-caged samples on each treatment were clipped, in April 1985 and 1986, to determine elk utilization by biomass.


A replicated, controlled study in 1986–1988 in a wooded area in Oklahoma, USA (Boggs et al. 1991) found that prescribed burning did not reduce bot fly Cuterebra infestation rates among rodents and cottontail rabbits Sylvilagus floridanus. The percentage of animals infested with Cuterebra larvae did not differ significantly between areas that were burned and sprayed with herbicide (14–20% of animals) and areas sprayed but not burned (18–20%). Eight plots (32.4 ha each) were burned annually in April, from 1985, and eight were not burned. Four burned and four unburned plots were sprayed with the herbicide, tebuthiuron (at 2.2 kg/ha), in March 1983. Remaining plots were treated with the herbicide, triclopyr (at 2.2 kg/ha), in June 1983. Rodents were surveyed using snap traps, in July–September and December–March 1986–1988. Cottontail rabbits were collected by shooting in January and July 1987–1988. Animal carcasses were opened up and examined for Cuterebra presence.


A replicated, controlled study in 1986–1988 of a forest and grassland site in Oklahoma, USA (Lochmiller et al. 1991) found that burning and spraying pastures with herbicide had mixed effects on cottontail rabbit Sylvilagus floridanus abundance when compared with spraying with herbicide alone. In seven of 10 comparisons, there was no significant difference between the number of cottontails found in pastures that were burned compared to those not burned. For three of 10 comparisons, there were more cottontails on burned areas (0.1–1.40 cottontails/ha) than on unburned areas (0–0.4). Eight 32-ha pastures were treated with the herbicides tebuthiuron or triclopyr (at 2.2 kg/ha in March 1983 or June 1983). Four of these pastures were burned in April 1985, 1986 and 1987. Rabbit density was estimated by walking transects, three times each July and February, from July 1986 to February 1988.


A replicated, controlled, paired sites study in 1991–1992 of scrubland in a national park in southern Spain (Moreno & Villafuerte 1995) found that burned plots hosted higher densities of European rabbits Oryctolagus cuniculus than did unburned plots. More rabbit pellets were counted in burned plots both in wet scrubland (burned: 11.6 pellets/plot/year; unburned: 9.8) and in dry scrubland (burned: 6.8; unburned: 1.6). Four plots each in wet and dry scrubland were burned in summer 1989. Each was paired with an unburned plot 1 km away, in the same habitat. Plots measured 300 × 200 m. Rabbit pellets were counted monthly in 1991 and 1992 at random sample locations in each plot.


A randomized, replicated, controlled study in 1988–1989, in a mixed forest and prairie site in Oklahoma, USA (Leslie Jr et al. 1996), found that burning areas of forest had mixed effects on use by white-tailed deer Odocoileus virginianus, depending on season and sex. Female deer preferred burned plots in spring and summer, but unburned plots in winter. Male deer preferred burned plots in summer and autumn. There was no habitat selection for other sex/season combinations. Data presented as proportions of radio-tracking locations. See paper for details. Four blocks, each containing five 32-ha plots, were studied. In each block, two plots were sprayed with herbicide and burned, two were sprayed with herbicide but not burned and one was not sprayed or burned. Burning was done each April in 1985–1987. Herbicide was applied in 1983. Ten female and seven male deer were radio-tracked, in 1988–1989, and the use of burned and unburned areas relative to their size was assessed.


A study in 1993–1995 of a prairie site in Oklahoma, USA (Coppedge & Shaw 1998) found that burned areas were selected for grazing by mixed age and sex bison Bison bison groups but were avoided by mature bull groups. Burned areas were selected in a higher proportion than their availability by mixed groups during 23% of observations and avoided during 13%. Unburned areas were selected in 0% of observations and avoided in 63%. Burned areas were selected by bull groups in 4% of observations, and avoided in 46%. Unburned areas were selected in 29% of observations, and avoided in 14%. Three hundred bison were reintroduced into a 1,973-ha study area in October 1993. The area was expanded by 292 ha in 1995. Controlled burns were carried out five times between September 1993 and December 1995. Bison usage of burned and unburned areas was surveyed 4–12 times/month in 1994 and 1995. Herds were generally mature (>5 year-old) bulls and mixed groups of cows, calves and younger bulls.


A replicated, randomized, controlled study in 1992–1994 of pine forest in a mountainous area of Arkansas, USA (Masters et al. 1998) found similar small mammal numbers and species richness after prescribed burning compared to in unburned forest. Small mammal capture rates in burned stands (animals caught on 2.3–7.1% of trap nights) did not significantly differ to those in unburned stands (3.9–7.4%). Average species richness did not differ between burned (2.7–4.3 species/plot) and unburned plots (1.7–4.7/plot). In nine plots (14–45 ha), mid-storey trees were removed and, the following spring, plots were burnt. In three plots, mid-storey trees were not removed and plots were not burned. Management was carried out to benefit red-cockaded woodpeckers Picoides borealis. Small mammals were surveyed using 80 live-trap stations/plot from 27 December to 4 January in 1992–1993 and 1993–1994.


A replicated, site comparison study in 1990–1996 of one prairie site in each of Nebraska and Oklahoma, USA (Biondini et al. 1999) found that plains bison Bison bison bison preferentially selected burned areas in most years. At one site, bison made more use of burned areas, relative to their availability, in five of seven years. There was no consistent pattern in the second site but, in the largest single range (65% of the site), bison selected burned areas in two of three years. In both cases, results were presented as deviation from modelled predictions. At one site, monitored from 1990–1996, approximately 13.5% of the site was burned each year. The second site, monitored from 1993–1996, had approximately 20% burned each year. Locations of mixed bison groups (females, bulls <4 years old and occasionally adult bulls) were determined at least monthly during respective monitoring periods at each site.


A before-and-after, site comparison study in 1986–1991 of a mixed grassland, shrubland and woodland site in Utah, USA (Smith et al. 1999) found that prescribed burning of sagebrush-grass shrublands and pinyon-juniper woodland increased use of these areas by Rocky Mountain bighorn sheep Ovis canadensis. Use of burned areas by sheep increased by 148% and use of unmanaged areas decreased by 45%. Following burning, more sheep used the area (82 sheep groups; average of 14 sheep/group) than before burning (117 sheep groups; average of nine sheep/group). On a 353-ha study area, 18% was burned and 49% was unmanaged. Additionally, 32% was clearcut (results not presented here). Sheep-use patterns were assessed pre-treatment, from June 1986 to September 1988, by observing 25–30 radio-collared sheep daily. Post-treatment use was assessed in June–September 1991, by counting sheep 62 times from an 11-km transect.


A before-and-after study in 2000 of a shrubland ranch in South Africa (Gureja & Owen-Smith 2002) found that prescribed burning of an area increased its use by roan antelope Hippotragus equinus and tsessebe Damaliscus lunatus but not by Lichtenstein’s hartebeest Alcelaphus lichtensteini or sable antelope Hippotragus niger. Roan were seen more frequently on burned areas (113 sightings) than on unburned areas (81 sightings) relative to their availability (31% of the study area was burned). Tsessebe showed a similar pattern (burned: 77 sightings; unburned: 54 sightings) as did zebra Equus burchelli (burned: 96 sightins; unburned: 24) There was no consistent selection for burned areas by hartebeest (burned: 27; unburned: 24) or sable antelope (burned: 12; unburned: 27). See paper for further details of timings of use of burned areas. Rare herbivores were farmed on a 2,700-ha game ranch. A 280-ha area was burned in October 2001 and a 565-ha area was burned in November 2001. Animal positions were surveyed from roads in early morning and late afternoon from October to December 2000.


A randomized, paired sites, before-and-after study in 2001–2002 of a shrubland site in Texas, USA (Rogers et al. 2004) found that burning plots already subject to mechanical vegetation clearance did not increase plot utilization by white-tailed deer Odocoileus virginianus relative to carrying out a second mechanical clearance. There was no significant difference in deer track counts between plots before (burning: 36; mechanical clearance: 37 track crossings/km) or after (burning: 43; mechanical clearance: 47 track crossings/km) treatments were applied. Ten plots (3–9 ha), established in a 6,154-ha study area, were paired by size, soil and vegetation. In March–April 1999, all plots were cleared of brush using a mechanical aerator pulled by a tractor. In September 2000, one plot from each pair was burned and the other was mechanically cleared a second time. Treatment assignment within pairs was random. Deer utilization was assessed by counting tracks along prepared track lanes, over three days, before and after treatments were applied.


A before-and-after, site comparison study in 1982–1997 in a shrubland in the Western Cape, South Africa (Watson et al. 2005) found that Cape mountain zebra Equus zebra zebra used burned areas more than unburned areas, and 92% of foals were produced in the three years post-fire compared to 8% in the three years pre-fire. Mountain zebras with access to burned areas used those areas 83% of the time (data not provided). By comparison, whilst the total areas burned were not stated, 23% of fires in the south east section and 89% of fires in the north burned ≤25% the area. Of the foals produced within three years of a fire, 24 were produced in the three years post-fire compared to two pre-fire. Mountain zebras were monitored in two of three sections of the 9,428-ha nature reserve, the north (2,263 ha) and south-east (3,583 ha), where zebras mostly occurred. One of nine fires recorded since establishment of the reserve in 1974 was a prescribed fire (year not stated), others were natural fires (average interval between fires was seven years). Use of burned and unburned areas was monitored between the fires of 1992 and 1996. The number of foals produced was monitored three years before and after the fires of 1982, 1992, 1996–1997.


A replicated, randomized, controlled, before-and-after study in 2001–2003 in North Carolina, USA (Greenberg et al. 2006) found that prescribed burning did not alter the abundance of eight small mammal species. After burning, the numbers of captures of eight small mammal species did not differ significantly between burned (0–28 animals/plot) and unburned plots (0–17 animals/plot). Similarly, before burning, numbers did not differ between plots assigned for burning (0–24 animals/plot) and unburned plots (0–19 animals/plot). See paper for full break-down of species abundances. Three blocks were established, containing plots of >14 ha. In each block, one plot was burned in March 2003 and one plot was not burned. Small mammals were live-trapped over 10 consecutive days and nights in July and August of 2001–2003.


A replicated, controlled, before-and-after study in 2002–2003 in a national park in North Dakota, USA (Milne-Laux & Sweitzer 2006) found that burning and clearing woody vegetation led to greater areas occupied by black-tailed prairie dog Cynomys ludovicianus colonies and more prairie dog burrows. The study does not distinguish between the effects of burning and mechanical vegetation clearance. After one year, prairie dog colonies occupied a greater area in plots where vegetation was cleared and burned (18–70% of available habitat) than in plots that were not cleared or burned (0–5%). Cleared and burned plots also had more new burrows (191–458) than did plots that were not cleared or burned (41–116). In each of three prairie dog colonies, a 2-ha plot just beyond the colony boundary underwent prescribed burning in May 2002 and mechanical brush removal in June 2002. Similar 2-ha plots that were not burned or cleared were used for comparison. Colony boundaries were mapped in May–September 2002 and May–August 2003. New burrows were mapped monthly during these periods.


A replicated, randomized, controlled, before-and-after study in 2001–2004 of a coniferous woodland in California, USA (Monroe & Converse 2006) found that prescribed fire did not increase the abundance of small mammals. Deer mouse Peromyscus maniculatus abundance was not significantly higher on burned than on unburned plots (results presented as modelled effect size). Similarly, lodgepole chipmunk Neotamias speciosus abundance and total small mammal biomass were not significantly higher in burned than in unburned plots. Nine plots, 15–20 ha in area, were studied. Three were burned between 28 September and 28 October 2001 and three were burned on 20 or 27 June 2002. Three plots were not burned. Treatments were allocated randomly to plots. Small mammals were sampled by live-trapping over eight consecutive nights and days each year. Sampling occurred in June–August 2001 (pre-treatment) and in June–September of 2002 and 2003 and June–August 2004.


A controlled study in 1984–1999 in a sagebrush shrubland in Montana, USA (Van Dyke & Darragh 2006) found that prescribed burning increased use of the area by elk Cervus canadensis. Elk used areas that had been burned more frequently (163–628 elk use days) than they used areas that had not been burned (32–298 elk use days). Burned areas had higher grass and forb cover and lower sagebrush cover than unburned areas. In October 1984, a 40-ha area of sagebrush shrubland was burned and, in April 1988, a 30-ha area was burned. Five permanent 404-m² plots (20.1 × 20.1 m) were established in each burned area and another five placed within the unburned portion, one in 1988 and four more in 1993. In June 1988–1993 and 1999, elk use of plots was estimated by counting the number of pellets within 1 m of six transects laid in each plot. Vegetation cover was estimated within five 25 × 51 cm randomly placed quadrats each plot. No livestock were present in the study area.


A replicated, randomized, controlled study in 2003–2004 in a mixed forest site in North Carolina, USA (Greenberg et al. 2007) found that prescribed fire did not alter abundances of four shrew species. In both sampling years, numbers of northern short-tailed shrews Blarina brevicaua caught did not differ between plots that were burned (2–6 animals/plot) and plots that were not burned (3–10 animals/plot). The same pattern was seen for smoky shrews Sorex fumeus (1–2 animals/plot vs 1–2 animals/plot), American pygmy shrews Sorex hoyi (2–4 animals/plot vs 0–2 animals/plot), and southeastern shrew Sorex longirostris (1–4 animals/plot vs 1–­5 animals/plot). In each of three blocks, established in 2001, one plot was burned in March 2003 and one plot was not burned. Plots were >14 ha. Shrews were surveyed using pitfall traps and drift fencing over 123 nights in 2003 and 125 nights in 2004.


A before-and-after, site comparison study in 1989–2001 within a mixed forest national park in Alberta, Canada (Hood et al. 2007) found that prescribed burning did not increase occupancy of lodges by beavers Castor canadensis. For lodges subject to prescribed burning once, the occupancy rate in the year after burning (25%) was lower than in the year before burning (41%). Some lodges were burned more than once and the odds of occupancy decreased by 58% for each additional burn. In a 194-km2 national park, occupancy of 734 beaver lodges, located between 1989 and 2001, was monitored by aerial or ground surveys, every 1–3 years. There were 121 prescribed fires (1–1,059 ha in extent) from 1979–2001. All but six (in October) were lit between April and June. Around 49% of the park was not burned in the study period.


A site comparison study in 1989–1999 in a sagebrush shrubland in Montana, USA (Van Dyke & Darragh 2007) found that prescribed burning was associated with a short-term, but not long-term, increase in elk Cervus canadensis usage. In the first year after burning, elk use of burned plots increased (from 116 to 210 elk use days) and declined on unburned plots (from 189 to 120 elk use days). After 10 years, elk use declined and was similar on both burned plots (72 elk use days) and unburned plots (56 elk use days). A 50-ha prescribed burn was made in April 1989, while 200 ha of the site was not burned. Five plots (404 m2 each) were established each in burned and unburned areas. Unburned plots, damaged by wildfire in 1991, were replaced in 1993 by three plots on remaining unburned land. Elk used the site from November–May. Elk pellet groups were counted in June 1989–1991, 1993, and 1999 along transects across each plot.


A replicated, randomized, controlled, before-and-after study in 2001–2003 of a forest in California, USA (Amacher et al. 2008) found that prescribed burning increased abundance of deer mouse Peromyscus maniculatus, but not California ground squirrels Spermophilus beecheyi, long-eared chipmunks Tamias quadrimaculatus or brush mice Peromyscus boylii. Deer mouse abundance increased with fire (after: 2.0; before: 0.5/100 trap nights) and declined at the same time in unburned plots (after: 1.3; before: 1.9/100 trap nights). Changes in capture rates from before to after treatments did not differ between burned and unburned plots for California ground squirrel, long-eared chipmunk or brush mouse (see paper for data). Forests stands were 14–29 ha each. Four stands were burned in October–November 2002 and four stands were not burned. Small mammals were live-trapped over nine consecutive days and nights in July–August 2001 (pre-burn) and 2003 (post-burn).


A replicated, controlled study in 2003–2004 of savanna grassland in a national park in Tanzania (Hassan et al. 2008) found that prescribed burning did not result in a higher level of herbivory by mammals. The amount consumed by herbivores varied by season but the overall average amount in burned plots (223 g/m2) was not significantly different to that in unburned plots (176 g/m2). Six study areas (each ≥10 ha, 1–40 km apart) were selected. Each consisted of one patch burned in May–July 2003 and one unburned patch. Herbivore consumption was measured monthly, from September 2003 to July 2004, as biomass differences between caged and uncaged areas in study plots.


A before-and-after study in 2003–2005 of grassland in Colorado, USA (Thompson et al. 2008), found that after a prescribed fire, swift foxes Vulpes velox denned more in the burned area but hunting use of the area did not significantly increase. Two foxes with core home ranges in the burn area denned inside the burn area more after the burn (100% of denning locations) than before (60–75% of locations). For four foxes with home ranges overlapping the burn area, the proportion of times they were located hunting in the burn area was not significantly higher after burning (45% of locations inside burn area) than before (32%). In January 2003–December 2004, ten foxes were radio-collared. Location was recorded ≥three times/week in 2003–2005. In March 2005, an area of 260 ha was burned by prescribed fire. Sufficient locations were obtained from four foxes to determine pre- and post-burn home range use.


A replicated, controlled study in 2005–2006 in a coniferous forest site in Oregon, USA (Long et al. 2009) found that thinning, followed by prescribed burning was associated with mixed effects on use by North American elk Cervus canadensis, depending on season, stand age and sex. Thinning and burning were carried out on the same plots, so their influences could not be separated. Female elk used plots burned two and three years previously, proportionally to their availability, preferentially selected 4-year-old burns, and avoided 5-year-old burns. Male elk spent less time in all burned plots relative to their availability (data presented as selection ratios). In 2001–2003, twenty-six forest stands (average 26 ha) were thinned between May and October, followed by prescribed fire during September or October of either the same or the following year. Twenty-seven similar stands (average 55 ha) were not thinned or burned. Radio-collars were fitted on 18 female and five male elk in spring 2005, and 30 female and nine male elk in spring 2006. Locations were recorded automatically, within 1 hour of sunset or sunrise.


A systematic review in 2008 of management aimed at restoring natural processes in conifer forests in southwestern USA (Kalies et al. 2010) found that, in forests where a low to moderate severity prescribed burn followed thinning, two mammal species showed positive responses (abundance or reproduction) compared to in unmanaged forests while three showed no response. Responses of tassel-eared squirrel Sciurus aberti and deer mouse Peromyscus maniculatus to burning after thinning were positive. No significant responses were detected for golden-mantled ground squirrel Spermophilus lateralis, gray-collared chipmunk Tamias cinereicollis or Mexican woodrat Neotoma mexicana. The specific effects of thinning versus burning were not separated, though a different part of the same study found no response of tassel-eared squirrel or deer mouse to thinning (without burning) by removal of small to intermediate diameter trees. The review used evidence from 22 studies and considered responses of species recorded in ≥5 studies. Responses of species to five ways of managing ponderosa pine Pinus ponderosa forests to recreate natural conditions and forest dynamics, and reduce wildfire risk, were assessed against responses to unmanaged controls.


A controlled study in 2004–2008 of heather moorland at a site in southern France (Cazau et al. 2011) found that burning heather (Calluna vulgaris and Erica tetralix) resulted in greater use of the moorland by mouflon Ovis gmelini musimon × Ovis sp. The average density of feeding mouflon (modelled to account for temperature-driven variations) was higher on burned plots (36/ha) than on unburned plots (5/ha). Before burning, each 360 × 80-m plot, had not been modified for >40 years. Two plots were burned in spring 2004 and two were left unburned. Mouflon use of plots was determined by counting feeding animals in each plot, at 20 minute intervals, for two hours up to sunset. In total, 668 such counts were made in 2004–2008.


A replicated, site comparison study in 2004–2010 of grassland in Western Australia, Australia (Legge et al. 2011) found that prescribed burns early in the dry season resulted in higher abundance and species richness of small mammals relative to extensive mid- to late-dry season wildfires. More mammals were found in plots with prescribed burning (5.7/plot) than in areas subject to wildfire (3.5/plot). The same was true for species richness (prescribed burning: 1.4/plot; wildfire: 1.1/plot). Fire history was determined from satellite imagery from 1999–2010. Prescribed burning was initiated in 2004. Areas burned less frequently than average were regarded as being managed by prescribed burning, earlier in the dry season. Areas burned more frequently than average were regarded as being wildfire areas, burned later in the dry season. Forty small mammal traps/0.25-ha plot were operated for 120 trap-nights/year. The number of plots surveyed is not stated.


A site comparison study in 2002–2003 in a shrubland site in Arizona, USA (Pasch & Koprowski 2011) found that prescribed burning resulted in smaller individual home ranges and shorter daily movements for Mexican fox squirrels Sciurus nayaritensis chiricahuae than did fire suppression. The average home range in prescribed burning areas (2.9 ha) was smaller than in fire suppression areas (6.6 ha). Average daily movements were lower in prescribed burning areas (212 m) than in fire suppression areas (336 m). In a 5,000-ha protected area, prescribed burning was initiated in 1976. In 1980–2001, there were 33 fires, over 260 ha total extent. Forty-three squirrels were live-trapped. Adults were radio-collared and data were analysed from 11 male and nine females, with ≥30 location fixes per season, from May 2002 to September 2003. Daily movements were measured by locating animals three times from 05:00 h to 11:00 h.


A site comparison study in 2007 of savanna grassland in a park in Tanzania (Eby & Ritchie 2013) found that vigilance (a measure of perceived predation risk) of Thomson’s gazelles Gazella thomsonii did not differ between those on burned and unburned areas. There was no difference between burned and unburned areas in group vigilance, individual vigilance or reaction time in presence of a model cheetah (data not presented). Gazelles were observed in July–August 2007 on 10 burned areas (burned after mid-April with 2 cm average new grass growth) and nine unburned grassland areas. Vigilance was defined as an animal raising its head above shoulder height. Group vigilance was the average proportion of individuals vigilant in a group at 5-minute intervals over one hour. Individual vigilance was recorded for randomly selected females, over 2 minutes. Reaction to a model cheetah was timed following model placement from a vehicle 60 m away from the group.


A before-and-after study in 2009–2010 on a shrubland reserve in South Africa (Isaacs et al. 2010) found that burning reduced the number of locations in which herbivores were present. In each of two main habitats, the proportion of locations at which impala Aepyceros melampus, kudu Tragelaphus strepsiceros, and zebra Equus burchelli were found was lower after burning than before. In one of two habitats, wildebeest Connochaetes taurinus and giraffes were present at a higher proportion of sites after burning than before burning (see paper for full details). Two habitat types were studied, based on underlying quartzite and sandy soils. Mammal presence was quantified by determining presence or absence of faecal pellets for each species in plots along transects through each habitat. Pellets were counted in April–May 2009, burns were carried out in June–November 2009 and plots were resampled in June 2010.


A replicated, site comparison study in 2006–2007 of scrubland at a site in Spain (Moreno & Rouco 2013) found more of some small mammal species at edges of old burned plots but not in plot centres or in younger plots, relative to unburned plots. In two of four comparisons, there were more Algerian mice Mus spretus in burned plots (64–109 captures/1,000 trap nights) than in unburned plots (32 captures/1,000 trap nights). For two of four comparisons there was no significant difference (burned: 8–22 captures/1,000 trap nights; unburned 32 captures/1,000 trap nights). In three of four comparisons, there was no difference in the abundance of wood mice Apodemus sylvaticus between burned (2–7 captures/1,000 trap nights) and unburned areas (2 captures/1,000 trap night). In one of four comparisons there were more wood mice in burned areas (burned: 14 captures/1,000 trap nights; unburned 2 captures/1,000 trap night). There was no significant differences in the abundance of greater white-toothed shrew Crocidura russula or garden dormouse Eliomys quercinus between burned and unburned plots. Three plots were burned in winter 2003 (three years before sampling), three plots were burned in winter 2006 (one year before sampling) and three were not burned. Plots covered 1 ha and were ≥1 km apart. Small mammals were surveyed by live-trapping in unburned plots and in centres and edges of burned plots, once each in summer, autumn, winter and spring from summer 2006 to spring 2007). Traps were operated for seven consecutive nights (and closed in the day).

Referenced papers

Please cite as:

Littlewood, N.A., Rocha, R., Smith, R.K., Martin, P.A., Lockhart, S.L., Schoonover, R.F., Wilman, E., Bladon, A.J., Sainsbury, K.A., Pimm S. and Sutherland, W.J. (2020) Terrestrial Mammal Conservation: Global Evidence for the Effects of Interventions for terrestrial mammals excluding bats and primates. Synopses of Conservation Evidence Series. University of Cambridge, Cambridge, UK.